Multi-scale variability in neuronal competition (1811.07012v1)

Abstract: We examine whether a single biophysical cortical circuit model can explain both spiking and perceptual variability. We consider perceptual rivalry, which provides a window into intrinsic neural processing since neural activity in some brain areas is correlated to the alternating perception rather than the constant ambiguous stimulus. The prevalent theory for spiking variability is a chaotic attractor called the balanced state; whereas, the source of perceptual variability is an open question. We present a dynamical model with a chaotic attractor that explains both spiking and perceptual variability and adheres to a broad set of strict experimental constraints. The model makes quantitative predictions for how both spiking and perceptual variability will change as the stimulus changes.